DZMM

How many mounts do you have? I just have one rclone team drive mount and 1 for unionfs - you only need to mount the team drive once. I ditched rc as it only helps populate the dir-cache a little bit faster after mounting, but it was causing problems with remounts that I couldn't be bothered to work out as the benefits are small. Yes, a 5-10s sleep after each mount helps avoid any problems. I'm doing ok with 4 my uploads to one team drive for about 40 hours so far - I've done way more than 750GB to the one TD, and more than 750GB/account so I think you're problems are probably because of your bad mounts.

sometimes I have a very rare error if I've been changing the mounts a lot, which you have in your initial setup. In plex, go to your library settings and browse (not add) to where one of your movies is e..g mount_unionfs/movies/avengers (2012???). If you can't browse to the folder in plex but you can in windows/krusader then restart the plex docker - for some reasons if you unmount/mount unionfs sometimes while plex, radarr, sonarr etc are running they can't see the folders. It's why in my script dockers are started after the mounts are running.

Plex behaves exactly the same as it would for a local folder - that's the whole point of the setup.

@Kaizac Thanks - I've moved all my files to one team drive with 4 uploads running for over a day now with no problems - each upload using a unique google account and different client credentials. My backlog will be gone today, which is great. For anyone else wanting to try this, this is how my rclone config looks. Because the TEAM_DRIVE_ID, PASSWORD_1 and PASSWORD_2 is the same all uploads go the same team drive and then I mount one of the vfs remotes, team_drive1_vfs, and add it to my unionfs mount: [team_drive1] type = drive client_id = client_id_google_1 client_secret = client_secret_google_1 scope = drive team_drive1 = TEAM_DRIVE_ID token = token1 [team_drive1_vfs] type = crypt remote = team_drive1:crypt filename_encryption = standard directory_name_encryption = true password = PASSWORD_1 password2 = PASSWORD_2 [team_drive2] type = drive client_id = client_id_google_2 client_secret = client_secret_google_2 scope = drive team_drive1 = TEAM_DRIVE_ID token = token2 [team_drive2_vfs] type = crypt remote = team_drive2:crypt filename_encryption = standard directory_name_encryption = true password = PASSWORD_1 password2 = PASSWORD_2 [team_drive3] type = drive client_id = client_id_google_3 client_secret = client_secret_google_3 scope = drive team_drive1 = TEAM_DRIVE_ID token = token3 [team_drive3_vfs] type = crypt remote = team_drive3:crypt filename_encryption = standard directory_name_encryption = true password = PASSWORD_1 password2 = PASSWORD_2 [team_drive4] type = drive client_id = client_id_google_4 client_secret = client_secret_google_4 scope = drive team_drive1 = TEAM_DRIVE_ID token = token4 [team_drive4_vfs] type = crypt remote = team_drive4:crypt filename_encryption = standard directory_name_encryption = true password = PASSWORD_1 password2 = PASSWORD_2

run scan library files in plex. In library settings also disable create video thumbnails, as plex will download the whole uploaded file. Many people say disable 'perform extensive media analysis' in Extras as it causes API bans, but I've not had any problems and because I run a local server I don't want to waste upload bandwidth if the Plex transcoding brain doesn't have all the data.

yes it is - you have to tell plex there's something new via 'Connect' in radarr or sonarr. Otherwise, Plex does a periodic scan on whatever frequency you've set

I think it was quite fast - just slow with the small files. After first run it does blocks so it's fast thereafter

5MB/s locally or 5MB/s upload to google? The upload to G is a normal rclone sync, so speed depends on the type of file e.g. lots of photos are slower

yep docker. Free version doesn't come with compression but it does block level backups so it's very efficient after the first backup.

I just use the free version to backup to my local folder that then gets encrypted when stored on google. Until last week I was using Crashplan for a few weeks, but it only managed to upload about 400GB in about 4 months....Duplicati I used for about 2 days and it started throwing up errors - made me lose confidence in it as a backup solution. CB did my local backup run really quickly that's now synced to gdrive. I did a quick fag packet calculation before I started switching to one teamdrive and I'm ok. I don't upload anything other than the media files - I'm not sure what value the nfo and srt files are to Plex etc, especially now that Plex's subtitles are getting better (99.99% of my content is in English anyway)??

lol that's insane. I might have to share servers with you to get some inspiration to beef up my library 🙂 @Spladge are you using multiple uploads to 1 or more teamdrives, or have you just been running one upload for a very long time? @Kaizac I'm moving a few TBs at a time and going very slow so I don't get duplicate folders. My slow method is to move from the old TD/gdrive to a 'move' folder on the destination teamdrive and then move the folders to the right folders within the mount using putty/windows

Thanks - I'm going to give one teamdrive another go.

thanks - I can't find the my.cnf file??

yes it does. You only need this if you are adding files to the mount on a different server to the one plex is running on e.g. @Spladge I believe has mounted rclone on a remote server and Plex is running on his local server, where he's also mounted rclone @Stupifier got it working - there's also plex_rcs

I've updated my scripts on github to fix a few issues people have been having, notably removing the rc command as I think it causes more problems than it's worth https://github.com/BinsonBuzz/unraid_rclone_mount

1. Plex treats gdrive files the same as it does local files, so it will use the same rules as to whether or not to direct play, transcode etc 2. as above - Plex does the work 3. No - if you have decent bandwidth you will hardly see a difference e.g. my cloud files launch in an average of about 5 seconds - some faster, some slower, which is a bit longer than it takes for a local disk to spinup

no - something must have gone wrong with your first mount - there normally shouldn't be anything in the unionfs folder before you mount

you've got something in mount_unionfs so the mount failed. Check what it is, delete and then run fusermount -uz /mnt/user/mount_unionfs/google_vfs unionfs -o cow,allow_other,direct_io,auto_cache,sync_read /mnt/user/rclone_upload/google_vfs=RW:/mnt/user/mount_rclone/google_vfs=RO /mnt/user/mount_unionfs/google_vfs in a temporary script

I don't quite understand some of the errors google throws out - you can't be over your upload limit as your mount is empty. As @nuhll says, it will fix itself soon. There's nothing to worry about as your usuing your own client_ID not the shared rclone one. your mount is working which is the main thing, so start uploading via the rclone_upload folder not by adding direct to the mount folder

yes. When unionfs 'deletes' a file from the RO directory, it doesn't actually delete the file it just hides it in the mount. What the script does is actually delete the files from the RO directory and then tidy up the hidden unionfs directory. For the RW directory it can delete the files, so the script doesn't need to do anything there.

That was for @nuhll (i) I think because you've got a mountcheck file in rclone_upload it's messing up the unionfs mount as it seeing two copies of the same file in two places. (ii) I'm not sure why the rclone_mount is empty though. Try this: 1. manually delete the encrypted mountcheck file on google drive and empty the trash just to be certain it's gone 2. delete the local file in rclone_upload 3. recreate the mountcheck file using touch in terminal 4. create a quick new temp script to unmount and remount Edit: remember to run script in the background #!/bin/bash fusermount -uz /mnt/user/mount_unionfs/google_vfs fusermount -uz /mnt/user/mount_rclone/google_vfs rclone mount --allow-other --buffer-size 512M --dir-cache-time 72h --drive-chunk-size 512M --fast-list --log-level INFO --vfs-read-chunk-size 128M --vfs-read-chunk-size-limit off gdrive_media_vfs: /mnt/user/mount_rclone/google_vfs & sleep 5 unionfs -o cow,allow_other,direct_io,auto_cache,sync_read /mnt/user/rclone_upload/google_vfs=RW:/mnt/user/mount_rclone/google_vfs=RO /mnt/user/mount_unionfs/google_vfs exit

Check my last reply - I updated it for you!

1. there's nothing in your crypt except your mountcheck file..... 2. you shouldn't be seein the mountcheck file in rclone_upload/google_vfs though as it's already in the cloud. What does your /mnt/user/mount_rclone/google_vfs look like?

Hmm. everything looks correct. Apologies if these questions seem obvious: 1. Are you definitely using the right password and password2? 2. the files that are already in gdrive are encrypted (goobledegook when you look in google)? 3. Are the google files in a folder called crypt on google i.e. not encrypted so you can see my drive/crypt when you look at Google?